Trisaccharides of Phenolic Glycolipids Confer Advantages to Pathogenic Mycobacteria through Manipulation of Host-Cell Pattern-Recognition Receptors

Article abstract of DOI:10.1021/acschembio.6b00568

Despite mycobacterial pathogens continue to be a threat to public health, the mechanisms that allow them to persist by modulating the host immune response are poorly understood. Among the factors suspected to play a role are phenolic glycolipids (PGLs), produced notably by the major pathogenic species such as Mycobacterium tuberculosis and Mycobacterium leprae. Here, we report an original strategy combining genetic reprogramming of the PGL pathway in Mycobacterium bovis BCG and chemical synthesis to examine whether sugar variations in the species-specific PGLs have an impact on pattern recognition receptors (PRRs) and the overall response of infected cells. We identified two distinct properties associated with the trisaccharide domains found in the PGLs from M. leprae and M. tuberculosis. First, the sugar moiety of PGL-1 from M. leprae is unique in its capacity to bind the lectin domain of complement receptor 3 (CR3) for efficient invasion of human macrophages. Second, the trisaccharide domain of the PGLs from M. tuberculosis and M. leprae share the capacity to inhibit Toll-like receptor 2 (TLR2)-triggered NF-κB activation, and thus the production of inflammatory cytokines. Consistently, PGL-1 was found to also bind isolated TLR2. By contrast, the simpler sugar domains of PGLs from M. bovis and Mycobacterium ulcerans did not exhibit such activities. In conclusion, the production of extended saccharide domains on PGLs dictates their recognition by host PRRs to enhance mycobacterial infectivity and subvert the host immune response.

Full text of DOI:10.1021/acschembio.6b00568

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Trisaccharides of Phenolic Glycolipids Confer
Advantages to Pathogenic Mycobacteria through
Manipulation of Host-Cell Pattern-Recognition Receptors
Ainhoa Arbues, Wladimir Malaga, Patricia Constant, Christophe
Guilhot, Jacques Prandi, and Catherine Astarie-Dequeker
ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.6b00568 • Publication Date (Web): 22 Aug 2016
Downloaded from http://pubs.acs.org on August 24, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 29
ACS Chemical Biology
1
2
3
4
5
Trisaccharides of Phenolic Glycolipids Confer Advantages to Pathogenic
Mycobacteria through Manipulation of Host-Cell Pattern-Recognition Receptors
6
7
8
9
Ainhoa Arbués a,b, Wladimir Malaga a,b, Patricia Constant a,b, Christophe Guilhot a,b, Jacques
Prandi a,b* and Catherine AstarieꢀDequeker a,b*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
CNRS, IPBS (Institut de Pharmacologie et de Biologie Structurale), BP 64182, 31077
Toulouse, France
b Université de Toulouse, UPS, 31077, Toulouse, France
supporting information
* Corresponding authors: C. AstarieꢀDequeker, CNRS, IPBS (Institut de Pharmacologie et de
Biologie Structurale), BP 64182, 205 route de Narbonne, Fꢀ31077 Toulouse Cedex 04,
France. Phone : 33ꢀ561 17 54 55; Fax : 33ꢀ561 17 59 94; eꢀmail: Catherine.Astarieꢀ
Dequeker@ipbs.fr; J. Prandi , CNRS, IPBS, BP 64182, 205 route de Narbonne, Fꢀ31077
Toulouse Cedex 04, France. Phone: 33ꢀ561 17 54 85; Fax: 33ꢀ561 17 59 94; eꢀmail:
Jacques.prandi@ipbs.fr.
Running title: Trisaccharides of PGLs: a weapon of pathogenic mycobacteria
Key words: Tuberculosis, leprosy, Mycobacterium, phenolic glycolipids, macrophages, immune
responses, CR3, TLR2.
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
ACS Paragon Plus Environment

Page 3 of 29
ACS Chemical Biology
1
2
3
Abstract
4
5
6
7
8
9
Despite mycobacterial pathogens continue to be a threat for public health, the mechanisms
that allow them to persist by modulating host immune response are poorly understood.
Among the factors suspected to play a role are phenolic glycolipids (PGLs), produced
notably by the major pathogenic species such as Mycobacterium tuberculosis and
Mycobacterium leprae. Here, we reported an original strategy combining genetic
reprogramming of the PGL pathway in Mycobacterium bovis BCG and chemical synthesis to
examine whether sugar variations in the speciesꢀspecific PGLs have an impact on pattern
recognition receptors (PRRs) and the overall response of infected cells. We identified two
distinct properties associated with the trisaccharide domains found in the PGLs from M.
leprae and M. tuberculosis. First, the sugar moiety of PGLꢀ1 from M. leprae is unique in its
capacity to bind the lectin domain of complement receptor 3 (CR3) for efficient invasion of
human macrophages. Second, the trisaccharide domain of the PGLs from M. tuberculosis
and M. leprae share the capacity to inhibit Tollꢀlike receptor 2 (TLR2)ꢀtriggered NFꢀκB
activation, and thus the production of inflammatory cytokines. Consistently, PGLꢀ1 was found
to also bind isolated TLR2. By contrast, the simpler sugar domains of PGLs from M. bovis
and Mycobacterium ulcerans did not exhibit such activities. In conclusion, the production of
extended saccharide domains on PGLs dictates their recognition by host PRRs to enhance
mycobacterial infectivity and subvert host immune response.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 29
1
2
3
4
5
6
7
8
9
Pathogenic mycobacteria synthesize unique complex lipids that, being positioned at the
outermost layer of their envelope, are able to interact with cells and serve as a tool to
manipulate host immune response. Among the factors suspected to play a role are phenolic
glycolipids (PGLs), which are produced by a limited number of mycobacterial species,
notably the major human pathogens Mycobacterium leprae, Mycobacterium tuberculosis and
Mycobacterium ulcerans –the aetiological agents of leprosy, tuberculosis and Buruli ulcer,
respectively. In the case of M. tuberculosis, only a few strains synthetize PGL and this was
shown to be associated with a hypervirulent phenotype in mice and rabbit (1, 2). Considering
that very little is known about the interaction of these lipids with the host cell, a better
understanding about their biological activity and molecular mechanism of action may
contribute to novel therapeutic strategies against mycobacterial pathogens.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
PGLs consist of a lipid core, a longꢀchain βꢀdiol diesterified by polymethylꢀbranched fatty
acids, which is well conserved with minor structural modifications (3). This core is ωꢀ
terminated by an aromatic nucleus that is glycosylated by a speciesꢀspecific oligosaccharide
(Figure 1). There is a large literature on the role of PGLs in the subversion of host immune
response during mycobacterial infection (for review, see (4)). However, their molecular
mechanisms of action are poorly known. Interestingly, their speciesꢀspecific nature suggests
that PGLs may be involved in functions unique to a given species. For instance, PGLꢀ1 from
M. leprae was reported to bind complement component C3 (5), with potential consequences
on the capacity of M. leprae to invade host cells, and to subsequently affect bacterial
resistance to intracellular killing by macrophages (6-8). PGLꢀ1 was also proposed to
modulate immune responses against M. leprae by suppressing the secretion of proꢀ
inflammatory cytokines by monocytes/macrophages (9, 10) and the activation and
proliferation of T lymphocytes (11-13). Yet, the latter effect is not restricted to PGLꢀ1, since
PGLs from several species inhibit nonꢀspecifically lymphoproliferative response to various
stimuli (12). Moreover, PGLs from M. marinum (PGLꢀmar) and M. tuberculosis (PGLꢀtb)
share the ability to abrogate the secretion of inflammatory cytokines (1, 14, 15), whereas
PGL from M. bovis (PGLꢀbov) and phthiocerol dimycocerosates (DIM), the common lipid
core, have no such effect (1). All together, these observations suggest that the biological
activities of PGLs are primarily mediated by their speciesꢀspecific saccharide domains. In line
with this, Elsaidi and coꢀworkers (16-18) have recently reported that subtle structural
differences in the glycan portion of synthetic PGL analogues modulate the inhibition of the
production of proꢀinflammatory cytokines triggered by the pattern recognition receptor (PRR),
Tollꢀlike receptor 2 (TLR2). We speculate that PGLs act as pathogenꢀassociated molecular
patterns (PAMPs) and that sensing of their sugar moiety by PRRs on innate immune cells
can trigger intracellular signalling cascades ultimately modulating the host microbicidal
3
ACS Paragon Plus Environment

Page 5 of 29
ACS Chemical Biology
1
2
3
4
5
response, as described for different types of glycosylated PAMPs such as βꢀglucan and
lipopolysaccharides (19).
We previously reported that reprogramming of the PGL pathway in the vaccine strain M.
bovis BCG to make it synthesize PGLꢀ1 endows this bacterium with a better capacity to
exploit complement receptor 3 (CR3) to invade human macrophages and subvert
inflammatory responses (10). CR3, also named Macꢀ1, CD11b/CD18 or αMβ2 integrin, is a
heterodimeric surface receptor expressed mainly on phagocytic cells that participates in cell
invasion, activation, chemotaxis and cytotoxicity (20). Interestingly, CR3 was also found to
function as a negative regulator of TLRꢀtriggered inflammatory responses (21, 22). This
receptor has the capacity to interact with a broad spectrum of ligands due to the presence of
multiple binding sites in its α chain. Notably, a lectinꢀlike domain that was characterized as a
binding site for βꢀglucans, and thus representing a potential recognition site for the
saccharide domain of PGLs (23).
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The present study was designed to get new insights into the molecular mechanisms
underlying PGL biological activities. To date, most studies addressing the role of PGLs
employed either wildꢀtype mycobacteria producing them or materials isolated from these
organisms. We focused on their speciesꢀspecific saccharide moieties using a crossꢀ
disciplinary approach, which combines their chemical synthesis and the genetic
reprogramming of the model strain M. bovis BCG to make it synthesize PGLs produced by
other mycobacterial pathogens in the context of a relevant mycobacterial envelope. We
provide evidence now that trisaccharide moieties of the PGLs from M. leprae and M.
tuberculosis act as PAMPs enabling these species to exploit host PRRs and consequently to
enhance infectivity and to favour evasion of host immune responses.
RESULTS
Construction of recombinant BCG strains producing various species-specific PGLs
In previous work, we have extensively characterized the speciesꢀspecific biosynthetic
pathways of PGLs in mycobacteria (10, 24). This enabled us to pursue the genetic
reprogramming of the PGL pathway in the model strain M. bovis BCG to make it synthesize
the speciesꢀspecific PGLs of other mycobacterial pathogens. BCG synthesizes a
monosaccharideꢀcontaining PGL (PGLꢀbov) because of several genetic defects (24).
Introduction of M. tuberculosis genes Rv1511 (gmdA) and Rv1512 (epiA), encoding proteins
responsible for the transformation of the GDPꢀDꢀmannose in GDPꢀLꢀfucose, and a functional
copy of Rv2958c, involved in the transfer of the second rhamnosyl residue, endows BCG
with the capacity of synthesize the trisaccharide domain of PGLꢀtb (Figure 1 and 2A).
Production of the PGLs from M. ulcerans (PGLꢀulc) and PGLꢀ1 requires first the disruption of
4
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 29
1
2
3
4
5
6
7
8
9
the Rv2959c ortholog, to avoid methylation of position 2 of the PGLꢀbov rhamnosyl residue.
We previously showed that simultaneous introduction of M. leprae genes ML0128 and
ML2348, encoding for glycosyltransferases, together with the methyltransferaseꢀencoding
genes ML0126, ML2346c and ML2347, leads to the production of the trisaccharide residue
characteristic of PGLꢀ1 (10). Based on protein similarities, we identified ML0126 as the gene
encoding likely the methyltransferase catalysing methylation of the position 3 of the first
rhamnosyl residue of PGLꢀ1 (10). Therefore, to engineer BCG to make it synthesize PGLꢀ
ulcꢀlike entities with a 3ꢀO-methylrhamnose sugar residue, we transferred an integrative
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
plasmid carrying the ML0126 gene into M. bovis BCG
ꢀRv2959c (Figure 1 and 2B). The
PGLs produced by the resulting recombinant strains were purified and the structure of their
saccharide domains was characterized by NMR spectroscopy (Figure S1). The NMR
analysis of PGL purified from rBCG ꢀRv2959::ML0126 supported that the structure of the
saccharide moiety corresponds to the expected 3-O-methylꢀαꢀLꢀrhamnoside found in PGLꢀulc
and this was confirmed by comparison with its synthetic oligosaccharide (Figure S1). A
replicative plasmid allowing the expression of gfp gene in mycobacteria was introduced in
BCG and the various PGLꢀproducing recombinant BCG (rBCG) strains in order to render
them fluorescent. This transfer had no impact on the PGL production (Figure 2A).
Chemical synthesis of PGL oligosaccharide domains
We also aimed to investigate the activities of the saccharide domain of PGL devoid of the
lipid core. For this purpose, we undertook to chemically synthesize these entities. The
structures of the targeted oligosaccharide epitopes of the PGL (OS-PGLs) from M. bovis, M.
ulcerans, M. leprae and M. tuberculosis are represented on Figure 1 (X = Me). All these
epitopes share common αꢀlinked Lꢀrhamnopyranoside units which might be 2ꢀOꢀmethylated
(PGLꢀbov and PGLꢀtb), 3ꢀOꢀmethylated (PGLꢀulc and PGLꢀ1) or 2,3ꢀdiꢀOꢀmethylated (PGLꢀ
1). They can also be 2ꢀOꢀglycosylated (PGLꢀ1), 3ꢀOꢀglycosylated (PGLꢀtb) or 4ꢀOꢀ
glycosylated (PGLꢀ1). Other speciesꢀspecific features of the epitopes are the terminal 3,6ꢀdiꢀ
OꢀmethylꢀDꢀglucosyl residue of PGLꢀ1 and the 2,3,4ꢀtriꢀOꢀmethylꢀLꢀfucosyl residue found in
PGLꢀtb.
Synthesis of the trisaccharidic epitope of PGLꢀ1 have been described by Brennan (25, 26),
Izumi (27) and more recently by the Lowary’s group (16). The PGLꢀtb epitope was first
prepared by Van Boom (28) and Fujiwara (29), and more recently by Scalan (30). The total
synthesis of the whole PGL of M. tuberculosis has been achieved by Minnaard’s group (31).
In our case, as all PGL epitopes were required, a convergent/divergent strategy was chosen
with the different Lꢀrhamnose units originating from a single advanced synthetic precursor.
Compound 2, the stannylene derivative of orthoester 1, was found ideally suited for this
purpose. After differential selective protections of the 3 and 4 positions of 2, acidic opening of
5
ACS Paragon Plus Environment

Page 7 of 29
ACS Chemical Biology
1
2
3
4
5
the orthoester ring will introduce an acetate group on position 2 to secure the
stereochemistry of all following glycosylations as α.
6
Preparation of the building blocks (Figure 3)
7
8
9
Differentiation of the 3 and 4 positions of 2 was obtained by selective methylation (or
benzylation) of the position 3 of stannylene 2 and later introduction on position 4 of a
protecting group, benzyl or silyl, orthogonal to the one previously put on position 3. This
strategy was straightforward and gave access to all needed rhamnosyl donors 3ꢀ6 in three to
five steps from orthoester 1. Finally, the last glycosyl units needed for the elaboration of all
OSꢀPGLs, 2,3,4ꢀtriꢀOꢀmethylꢀLꢀfucosyl thioglycoside 7 and 3,6ꢀdiꢀOꢀmethylꢀDꢀglucosyl
imidate 8 were prepared by slight modifications of reported procedures (32, 33).
Stannylene 2 was prepared from diol 1 and one equivalent of dibutyltinoxide in refluxing
toluene with elimination of water. Selective 3ꢀOꢀbenzylation of 2 was carried out under
standard conditions with benzyl bromide and tetrabutylammonium bromide in
dimethylformamide (DMF) at 40°C (34). After workup, the crude product was silylated with
tertꢀbutyldimethylsilyl chloride (TBDMSCl) in pyridine to give orthoester 9 in 58% isolated
yield for the two steps. Introduction of the anomeric trichloroacetimidate was done in two
steps, acidic hydrolysis of the orthoester was followed by treatment of the hemiacetal with
trichloroacetonitrile and 1,8ꢀdiazabicyclo[5.4.0]undecꢀ7ꢀene (DBU) in dichloromethane
(DCM). Compound 5 was isolated in 90% yield after these two steps. The very same strategy
was used for preparation of donors 3 and 4. Stannylene 2 was first selectively methylated on
the 3 position using methyl iodide and tetrabutylammonium bromide in DMF at 40°C.
However, the intermediate 3ꢀOꢀmethyl orthoester 10 was found to be highly volatile,
excluding its isolation, even in a crude form. The reaction mixture was thus filtered on silica
(THF elution) and THF was evaporated in the cold. The resulting solution of 10 in DMF was
directly used for benzylation (excess NaH, BnBr, 0°C) to give known orthoester 11 in a 72%
yield for the two steps (35). In a similar manner, treatment of the DMF solution of crude 10
with an excess of TBDMSCl in presence of imidazole gave the 4ꢀOꢀsilyl orthoester 12 in 64%
yield from 2. The two orthoesters 11 and 12 were then transformed to the corresponding
trichloroacetimidates according to the sequence use above for 5, known imidates 3 (36) and
4 (37) were efficiently obtained in respective yields of 77% and 68%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The last rhamnosyl donor 6 was obtained in three steps from orthoester 1. Benzylation of the
3 and 4 positions of 1 afforded dibenzyl orthoester 13 in 90% isolated yield (38). Introduction
of the activable anomeric imidate was done as above and donor 6 was obtained in 77% from
13 (39).
Elaboration of the epitopes of PGLs (Figure 4)
OS-PGL-bov. The synthesis of this epitope started from imidate 6. Glycosylation of a slight
excess of pꢀcresol with 6 (cat. TMSOTf, DCM, –20°C) gave αꢀglycoside 14 in 73% yield. The
6
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 29
1
2
3
4
5
6
7
8
acetate group of 14 was removed under basic methanolic conditions, and alcohol 15 was
then methylated with sodium hydride and methyl iodide in THF to give 16. Overall yield for
these two steps was 65%. Deprotection of 16 by hydrogenolysis (H2, Pd(OH)2/C) gave OS-
PGL-bov in 95% yield.
9
OS-PGL-ulc and OS-PGL-1. The synthesis of OS-PGL-ulc and OS-PGL-1 shares the same
two first steps. Glycosylation (cat. TMSOTf, DCM, –20°C) of pꢀcresol with rhamnosyl donor 3
gave αꢀrhamnoside 17 in 72% yield and removal of the acetate group of 17 gave alcohol 18
in 90%. From 18, OS-PGL-ulc was obtained in quantitative yield after hydrogenolysis (H2,
Pd(OH)2/C) of the benzyl protecting groups.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the elaboration of OS-PGL-1, 18 was glycosylated with 4 and gave αꢀdirhamnoside 19 in
75% yield. Introduction of the 2’ꢀOꢀmethyl group on 19 was done in two steps as described
above for the preparation of OS-PGL-bov and 21 was obtained in 78% yield from 19.
Deprotection of the 4’ꢀOꢀsilyl group of 21 was done with tetrabutylammonium fluoride
(NBu4F) in THF and alcohol 22 was isolated in 85% yield. Final glycosylation of 22 with
glucosyl donor 8 gave trisaccharide 23 in 91% yield as a single βꢀanomer on the new
anomeric center. Deprotection of 23 to OS-PGL-1 was done in two steps (MeONa, MeOH,
then H2, Pd(OH)2/C, AcOEt/MeOH) and OS-PGL-1 was isolated in 65%.
OS-PGL-tb. As the two Lꢀrhamnopyranoside units of the PGLꢀtb are glycosylated on position
3, only one rhamnose building block 5 was needed for the synthesis of this epitope.
Glycosylation of pꢀcresol with 5 gave pure αꢀLꢀrhamnopyranoside 24 in 68% yield.
Introduction of the 2ꢀOꢀmethyl group on 24 was done in two steps as described above for
OS-PGL-bov and 26 was obtained in 78% yield. Selective unmasking of the 3 position of 26
was carried out by hydrogenolysis and gave alcohol 27 in 93% yield. Glycosylation of 27 with
imidate 5 gave αꢀdirhamnoside 28 in 89% yield and the 3’ꢀOꢀbenzyl protecting group of the
nonꢀreducing rhamnose unit was removed by catalytic hydrogenation to give alcohol 29 in
95% yield. Glycosylation of this disaccharide with thiofucoside 7 (Nꢀiodosuccinimide, cat.
TMSOTf, CH2Cl2, –20 °C to room temperature) gave trisaccharide 30 in good yield (70%).
Final deprotection was carried out by treatment of 30 with sodium methanolate in methanol
followed by reaction with NBu4F in THF. The requested epitope OS-PGL-tb was isolated in
67% yield. All physicoꢀchemical data for new compounds are described in the Supporting
Information.
PGL-1 from M. leprae is unique to confer BCG with increased capacity to exploit CR3
lectin domain
Once the microbiological and molecular tools were generated, we undertook the comparison
of the biological activities of the various PGLs
.
We have previously demonstrated that PGLꢀ1
7
ACS Paragon Plus Environment

Page 9 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
production endows rBCG with increased capacity to infect human macrophages (hMDM)
(10). Therefore, we examined the role of other speciesꢀspecific PGLs in this phenotype. After
2h of contact under nonꢀopsonic conditions, all the strains were found to be efficiently
phagocytosed (Figure 5A). Comparative analysis showed that rBCG expressing PGLꢀtb and
PGLꢀulc infected cells to the same extent as BCG, but we observed a marked increase in the
percentage of hMDM infected by rBCGꢀPGLꢀ1 (Figure 5A). These data confirm that
expression of PGLꢀ1 specifically enhances invasivity of rBCG and argue in favour of a critical
role of the unique trisaccharide residue of PGLꢀ1 in this phagocytosis process.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Next, we assessed whether the lectin domain of CR3 was involved in the PGLꢀ1ꢀmediated
phagocytosis. hMDM were preꢀtreated with blocking antibodies (Ab) raised against either an
undetermined extracellular region (2LPM19c) or the lectin domain (VIM12) of human CR3,
and their impact on the differential uptake of BCG and rBCGꢀPGLꢀ1 was subsequently
evaluated (Figure 5B). In agreement with our previous results, none of the two Ab was able
to affect the phagocytosis of BCG (10). By contrast, both of them significantly decreased the
uptake of rBCGꢀPGLꢀ1 in a specific manner, since no effect was observed with the isotype
control. Interestingly, blockade of CR3 lectin domain restored phagocytosis of rBCGꢀPGLꢀ1
at a level comparable to that observed with BCG (Figure 5B). Taken together, these results
strongly support that PGLꢀ1, through its speciesꢀspecific sugar moiety, confers to BCG the
capacity to exploit the lectin site of CR3 for a more efficient invasion of macrophages.
Remarkably, this capacity was not shared by the other speciesꢀspecific PGLs.
Lyn is critical in PGL-1-mediated increase of phagocytosis by hMDM
To our knowledge, the signalling process of the lectin domain of CR3 is poorly known but it is
well known that the tyrosine kinases of the Src and Syk families contribute to initiate the
integrin signalling pathway (for review, see (40)). In addition, Syk and the Src kinase Lyn
have been shown to be required for the complementꢀmediated phagocytosis of pathogens
(22, 41). To test whether these kinases are involved in the differential uptake of BCG and
rBCGꢀPGLꢀ1, we selectively knocked them down by siRNAꢀmediated gene silencing (42)
(Figure 6A). When compared to siRNA control, Lyn and Syk knockdown did not impair the
uptake of BCG by hMDMs (Figure 6B). By contrast, Lyn, but not Syk, inactivation greatly
altered rBCGꢀPGLꢀ1 uptake, providing evidence that only Lyn is required for the increase of
CR3ꢀmediated phagocytosis induced by PGLꢀ1.
The unique sugar moiety in PGL-1 confers its capacity to specifically bind CR3
Our results of the phagocytosis experiments support the recognition of PGLꢀ1 through the
lectin domain of CR3. However, evidence demonstrating the direct binding were still missing.
We developed a solid phase assay in which increasing concentrations (0.5–25 ꢁM) of
8
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 29
1
2
3
4
5
6
7
8
9
isolated PGLꢀ1 were incubated with immobilized human CR3. Bound PGLꢀ1 was detected
using an Ab specific for PGLꢀ1 saccharide domain. Using this assay, we showed that native
PGLꢀ1, purified from M. lepraeꢀinfected armadillos, specifically binds human CR3 in a doseꢀ
dependent manner (Figure 7A). By contrast, no binding was detected when a nonꢀrelevant
protein was coated into the plate wells (No CR3). Next, we wondered if PGLꢀ1 sugar moiety
was sufficient for CR3 recognition. Immobilized CR3 was incubated with either 5 ꢁM purified
PGLꢀ1 or 50 ꢁM OSꢀPGLꢀ1. Detection using antiꢀPGLꢀ1 Ab showed that both molecules
were able to bind CR3 (Figure 7B). However, the binding observed for OSꢀPGLꢀ1 was lower
than for native PGLꢀ1, even when used at a tenꢀfold higher concentration, suggesting that
the lipid core enhances the affinity of CR3 for the PGLꢀ1 saccharide epitope.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Since PGLꢀ1 was the only PGL inducing an increase in the CR3ꢀdependent uptake of rBCG
by hMDM, we asked whether the direct binding to CR3 was also specific of PGLꢀ1 sugar
moiety. CR3 is a remarkably versatile receptor recognizing both endogenous ligands as well
as microbial molecules due to the presence of two binding domains (20). The binding of nonꢀ
protein ligands to CR3 is thought to be mediated by its lectin domain, which recognizes
several sugars including yeast β(1ꢀ3)ꢀglucans at highest affinity (23). Of note, the specific
terminal disaccharide of PGLꢀ1 contains a β(1ꢀ3)ꢀlinked glucose that may be crucial for the
interaction with CR3 lectin domain. Since specific Abs for each PGL were not available, we
developed a competition assay to address this question. CR3 was preꢀincubated with the
various synthetic OSꢀPGLs (50 ꢁM) before being added to wells coated with purified PGLꢀ1.
Bound CR3 was subsequently detected using an Ab raised against CD11b. As predicted, the
binding of CR3 to coated PGLꢀ1 was dramatically reduced only in the presence of the
synthetic PGLꢀ1 saccharide moiety (Figure 7C). Taken together, these results clearly
established that the specific interaction with CR3 is mediated by direct binding of the
trisaccharide domain to this receptor.
Synthetic epitopes of PGL-1 and PGL-tb inhibit TLR2-dependent NF-κB activation
Next, we investigated the molecular mechanisms involved in the immunomodulatory
properties of some of the PGLs produced by mycobacterial pathogens. We previously
reported that PGLꢀ1 production endows rBCG with a higher capacity to dampen inflammatory
responses in infected macrophages (10). This prompted us to examine the role of other
speciesꢀspecific PGLs. First, we used a NFꢀκBꢀreporter THPꢀ1 cell line designed to monitor
activation of the nuclear factor Kappa B (NFꢀκB), a transcription factor controlling the
expression of multiple inflammatory genes. After 16h of infection, BCG induced a strong
increase of the mean Abs630nm value indicative of expression of the reporter gene. In
comparison with BCG, while infection with rBCGꢀPGLꢀulc induced a similar response, the
9
ACS Paragon Plus Environment

Page 11 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
response induced by rBCGꢀPGLꢀtb or rBCGꢀPGLꢀ1 tended to be lower or significantly
diminished (Figure 8A). We further monitored the level of the proꢀinflammatory cytokine TNFꢀ
α in the supernatant of hMDM infected with BCG compared to that of rBCG expressing PGLꢀ
1 or PGLꢀtb. After 2h of infection, we confirmed that rBCGꢀPGLꢀ1 induced less secretion of
TNFꢀα than BCG (Figure 8B), despite the fact that it is more efficiently internalized (Figure
5A) (10). Similarly, infection with rBCGꢀPGLꢀtb led to relatively lower TNFꢀα secretion when
compared to BCG (Figure 8B). These results indicate that PGLꢀ1 and, to a lesser extent,
PGLꢀtb share the capacity to dampen the secretion of TNFꢀα, in agreement with previous
studies (1, 10), and strongly support the hypothesis that the immunosuppressive effects of
PGLs require trisaccharide moieties (18).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To further explore this hypothesis, we tested the effect of purified PGLs and synthetic PGL
epitopes in NFꢀkB activation. None of the tested compounds have an effect by themselves
(data not shown). Recently, Elsaidi and coꢀworkers established that truncated analogues of
the PGLs produced by M. leprae and M. tuberculosis, but not M. bovis, inhibit the release of
cytokines induced by TLR2 (16, 17), a PRR that promotes the synthesis of proꢀinflammatory
cytokines through the activation of NFꢀκB. This prompted us to evaluate the impact of
isolated compounds on cells simultaneously treated with a TLR2 agonist, Pam3CSK4. As
expected, this positive control induced a marked activation of NFꢀκB (Figure 8C). Addition of
5 ꢁM purified PGLꢀ1 dramatically decreased Pam3CSK4ꢀinduced activation of NFꢀκB,
whereas 10 ꢁM DIM, the common lipid core, had no such effect (Figure 8C). When 50 ꢁM
synthetic OSꢀPGLs were tested in the same system, we found that only the trisaccharide
moieties of PGLꢀtb and PGLꢀ1 were able to inhibit TLR2ꢀdependent NFꢀκB activation (Figure
8C), and thus supporting the results published by Elsaidi and coꢀworkers (16-18). However,
since hMDM and THPꢀ1 cells express a large repertoire of receptors, we cannot distinguish if
trisaccharideꢀcontaining PGLs exert their inhibitory effect via a direct recognition by TLR2, or
through potential crosstalk with another receptor recognizing the PGLs. Thus, we decided to
switch to a HEK reporter cell line expressing only human TLR2. Cells were treated for 16h
with 10 mg mlꢀ1 Pam3CSK4 in addition to purified DIM (10 ꢁM), purified PGLꢀ1 (5 ꢁM), or the
synthetic OSꢀPGLs (50 ꢁM). Similar to the results obtained in THPꢀ1 cells, native PGLꢀ1 and
saccharide epitopes of PGLꢀtb and PGLꢀ1 were found to inhibit NFꢀκB activation triggered by
Pam3CSK4 (Figure 8D). This observation led us to propose that recognition by TLR2
mediates, at least partially, the inhibition of NFꢀκB activation induced by trisaccharideꢀ
containing PGLs. In line with this, purified PGLꢀ1 was found to also bind immobilized human
TLR2 in a solid phase assay (Figure 8E). Collectively, these results evidence that PGLs from
M. leprae and M. tuberculosis can act as ligands of TLR2 to antagonize the signalling
pathway downstream this receptor, and thus to dampen TNFꢀα secretion. Furthermore, we
10
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 29
1
2
3
4
5
6
7
8
9
showed that the downꢀregulation of TNFꢀα secretion is mediated in part by the trisaccharide,
but not the monosaccharide epitopes of PGLs, through an inhibition of TLR2ꢀdependent NFꢀ
κB activation. This is consistent with the capacity of both PGLꢀtb and PGLꢀ1 to inhibit the
secretion of monocyte chemoattractant proteinꢀ1 (MCPꢀ1/CCL2) by macrophages (1, 16),
unlike the effect induced by the PGLꢀmar, a PGLꢀulcꢀlike, in the zebrafish model that is TLRꢀ
independent (43). Despite the fact that DIM alone showed no activity, native PGLꢀ1 was a
tenꢀtimes more potent inhibitor than its corresponding saccharide epitope. These results
suggest that the sugar domain is responsible for the specificity of TLR2 for some of the
PGLs, which is determined by the length of the oligosaccharide; whereas the common lipid
core would rather enhance the affinity of the receptor for the trisaccharide epitope, for
example, by improving its presentation. This could be explained by the presence of several
binding sites in TLR2 involved in the recognition of PGL; for instance, one sensitive to the
length of the saccharide domain and that confers the speciesꢀspecificity, and another one
with a hydrophobic pocket to accommodate the lipid chains present in both whole PGL and
lipopeptides–TLR2 favourite ligands.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We previously described a critical role of CR3 in the down modulation of TNFꢀα secretion
induced by rBCGꢀPGLꢀ1 when compared to BCG (10). While little is known about how the
engagement of CR3 by a pathogen can mediate regulation of TLRꢀdependent signaling
pathways, a recent study shows that TLR2ꢀdependent proꢀinflammatory responses can be
down modulated by a negative feedback, which engages CR3 and a downstream signaling
pathway including Lyn kinase (22, 41). To assess the role of Lyn in mediating immune
suppression pathway(s) during the phagocytosis of rBCGꢀPGLꢀ1, we selectively inactivated it
by siRNAꢀmediated gene silencing and quantified TNFꢀα secretion. Lyn knockdown
decreased TNFꢀα secretion by both BCG and rBCGꢀPGLꢀ1ꢀinfected cells (Figure 8F).
Nevertheless, this had no impact on the capacity of rBCGꢀPGLꢀ1 to dampen TNFꢀα
production in comparison with BCG (Figure 8F). Thus, under these conditions, engagement
of Lyn in the phagocytic process did not contribute to CR3ꢀmediated immune suppression
and crosstalk with the TLR2 signaling pathway (22). Whether PGLꢀ1 expression engages a
crosstalk between CR3 and TLR2 for the modulation of proꢀinflammatory cytokines remains
an open question, but we clearly demonstrated that PGLꢀ1 dampens TLR2 signalling
independently of its capacity to engage CR3 in bacterial uptake.
This study demonstrates that trisaccharide moieties of PGL from M. leprae and M.
tuberculosis act as PAMPs to enable bacteria to exploit host PRRs and subsequently
modulate the host immune response. Their surface expression enhances pathogen infectivity
through recognition of the lectin site of CR3 and also inhibits TLR2ꢀdependent signaling
cascades, by directly binding TLR2, that ultimately favors the resilience of M. tuberculosis
11
ACS Paragon Plus Environment

Page 13 of 29
ACS Chemical Biology
1
2
3
4
5
and M. leprae to subvert the hostile environment pose by the host cell. This report adds to
the growing body of data indicating that M. tuberculosis is equipped with surface molecules
aimed to inhibit TLRꢀorchestrating innate defences.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
METHODS
An extended methods section describing assays for lipids purification, synthesis of PGL
epitope, microbiology (bacterial growth conditions, biochemical analysis) and cell biology
(culture conditions, reagents, siRNA transfection, quantification of NFꢀκB activity and TNFꢀα
secretion and Western blot studies) can be found in the Methods section of the Supporting
Information.
ACKNOWLEDGEMENTS
We are grateful to L. Blanc, C. Robert and J. Nigou (IPBS, Toulouse, France) for HEK and
THPꢀ1 experiments, to A. Benard (IPBS, Toulouse, France) for his help with ELISA, and G.
LugoꢀVillarino (IPBS, Toulouse, France) for his contribution to siRNAꢀmediated gene
silencing and proofꢀreading of this manuscript. PGLꢀ1 purified from M. leprae as well as
antibodies against PGLꢀ1 were provided by BEI resources. We thank the Genotoul
plateforms TRI (IPBS, Toulouse, France) for the use of epifluorescence microscope. We
thank A. Lemassu and M. Daffé (IPBS, Toulouse, France) for access to the various
apparatus and the help with the structural characterization of glycolipids. AA is recipient of a
fellowship from a European Research Grant of Marie SkłodowskaꢀCurie actions (PIEFꢀGAꢀ
2012ꢀ329818). This work was supported by the Centre National de la Recherche Scientifique
(CNRS), the National Research Agency (ANRꢀ11ꢀBSV3ꢀ001)) and by European Regional
Development Funds.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
AUTHORS CONTRIBUTION
Author contributions: AA designed and conducted in vitro and ex-vivo experiments and
analyzed the results. WM carried out constructions of plasmid and recombinant strains. JP
designed experiments and performed chemical synthesis. PC performed structural
biochemistry experiments. CAD designed experiments and performed some exꢀvivo
experiments and analyzed data. CG and CAD conceived the project and obtained financial
12
ACS Paragon Plus Environment

ACS Chemical Biology
Page 14 of 29
1
2
3
4
5
support from ANR. AA, JP, CG and CAD wrote the paper. All authors reviewed the results
and approved the final version of the manuscript.
6
7
8
REFERENCES
9
1.
Reed, M. B., Domenech, P., Manca, C., Su, H., Barczak, A. K., Kreiswirth, B. N.,
Kaplan, G., and Barry, C. E., 3rd. (2004) A glycolipid of hypervirulent tuberculosis
strains that inhibits the innate immune response, Nature 431, 84ꢀ87.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.
Tsenova, L., Ellison, E., Harbacheuski, R., Moreira, A. L., Kurepina, N., Reed, M. B.,
Mathema, B., Barry, C. E., 3rd, and Kaplan, G. (2005) Virulence of selected
Mycobacterium tuberculosis clinical isolates in the rabbit model of meningitis is
dependent on phenolic glycolipid produced by the bacilli, J Infect Dis 192, 98ꢀ106.
Daffe, M., and Laneelle, M. A. (1988) Distribution of phthiocerol diester, phenolic
mycosides and related compounds in mycobacteria, J Gen Microbiol 134, 2049ꢀ2055.
Neyrolles, O., and Guilhot, C. (2011) Recent advances in deciphering the contribution
of Mycobacterium tuberculosis lipids to pathogenesis, Tuberculosis (Edinb) 91, 187ꢀ
195.
3.
4.
5.
Schlesinger, L. S., and Horwitz, M. A. (1991) Phenolic glycolipidꢀ1 of Mycobacterium
leprae binds complement component C3 in serum and mediates phagocytosis by
human monocytes, J Exp Med 174, 1031ꢀ1038.
6.
Neill, M. A., and Klebanoff, S. J. (1988) The effect of phenolic glycolipidꢀ1 from
Mycobacterium leprae on the antimicrobial activity of human macrophages, J Exp
Med 167, 30ꢀ42.
7.
Chan, J., Fujiwara, T., Brennan, P., McNeil, M., Turco, S. J., Sibille, J. C., Snapper,
M., Aisen, P., and Bloom, B. R. (1989) Microbial glycolipids: possible virulence factors
that scavenge oxygen radicals, Proc Natl Acad Sci U S A 86, 2453ꢀ2457.
Launois, P., Blum, L., Dieye, A., Millan, J., Sarthou, J. L., and Bach, M. A. (1989)
Phenolic glycolipidꢀ1 from M. leprae inhibits oxygen free radical production by human
mononuclear cells, Res Immunol 140, 847ꢀ855.
8.
9.
Silva, C. L., Faccioli, L. H., and Foss, N. T. (1993) Suppression of human monocyte
cytokine release by phenolic glycolipidꢀI of Mycobacterium leprae, Int J Lep Other
Mycobact Dis 61, 107ꢀ108.
10.
Tabouret, G., AstarieꢀDequeker, C., Demangel, C., Malaga, W., Constant, P., Ray, A.,
Honore, N., Bello, N. F., Perez, E., Daffe, M., and Guilhot, C. (2010) Mycobacterium
leprae phenolglycolipidꢀ1 expressed by engineered M. bovis BCG modulates early
interaction with human phagocytes, PLoS Pathog 6, e1001159.
11.
Mehra, V., Brennan, P. J., Rada, E., Convit, J., and Bloom, B. R. (1984) Lymphocyte
suppression in leprosy induced by unique M. leprae glycolipid, Nature 308, 194ꢀ196.
13
ACS Paragon Plus Environment

Page 15 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
12.
13.
Fournie, J. J., Adams, E., Mullins, R. J., and Basten, A. (1989) Inhibition of human
lymphoproliferative responses by mycobacterial phenolic glycolipids, Infect Immun 57,
3653ꢀ3659.
Dagur, P. K., Sharma, B., Upadhyay, R., Dua, B., Rizvi, A., Khan, N. A., Katoch, V.
M., Sengupta, U., and Joshi, B. (2012) Phenolicꢀglycolipidꢀ1 and lipoarabinomannan
preferentially modulate TCRꢀ and CD28ꢀtriggered proximal biochemical events,
leading to Tꢀcell unresponsiveness in mycobacterial diseases, Lipids Health Dis 11,
119.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
14.
Sinsimer, D., Huet, G., Manca, C., Tsenova, L., Koo, M. S., Kurepina, N., Kana, B.,
Mathema, B., Marras, S. A., Kreiswirth, B. N., Guilhot, C., and Kaplan, G. (2008) The
phenolic glycolipid of Mycobacterium tuberculosis differentially modulates the early
host cytokine response but does not in itself confer hypervirulence, Infect Immun 76,
3027ꢀ3036.
15.
16.
17.
18.
19.
Robinson, N., Kolter, T., Wolke, M., Rybniker, J., Hartmann, P., and Plum, G. (2008)
Mycobacterial phenolic glycolipid inhibits phagosome maturation and subverts the
proꢀinflammatory cytokine response, Traffic 9, 1936ꢀ1947.
Elsaidi, H. R., Barreda, D. R., Cairo, C. W., and Lowary, T. L. (2013) Mycobacterial
phenolic glycolipids with a simplified lipid aglycone modulate cytokine levels through
Tollꢀlike receptor 2, Chembiochem 14, 2153ꢀ2159.
Elsaidi, H. R., and Lowary, T. L. (2014) Inhibition of cytokine release by
mycobacterium tuberculosis phenolic glycolipid analogues, Chembiochem 15, 1176ꢀ
1182.
Elsaidi, H. R., and Lowary, T. L. (2015) Effect of phenolic glycolipids from
Mycobacterium kansasii on proinflammatory cytokine release. A structureꢀactivity
relationship study, Chem Sci 6, 3161ꢀ3172.
Mahla, R. S., Reddy, M. C., Prasad, D. V., and Kumar, H. (2013) Sweeten PAMPs:
Role of Sugar Complexed PAMPs in Innate Immunity and Vaccine Biology, Front
Immunol 4, 248.
20.
21.
Ehlers, M. R. (2000) CR3: a general purpose adhesionꢀrecognition receptor essential
for innate immunity, Microbes Infect 2, 289ꢀ294.
Han, C., Jin, J., Xu, S., Liu, H., Li, N., and Cao, X. (2010) Integrin CD11b negatively
regulates TLRꢀtriggered inflammatory responses by activating Syk and promoting
degradation of MyD88 and TRIF via Cblꢀb, Nature Immunol 11, 734ꢀ742.
Dai, S., Rajaram, M. V., Curry, H. M., Leander, R., and Schlesinger, L. S. (2013) Fine
tuning inflammation at the front door: macrophage complement receptor 3ꢀmediates
phagocytosis and immune suppression for Francisella tularensis, PLoS Pathog 9,
e1003114.
22.
14
ACS Paragon Plus Environment

ACS Chemical Biology
Page 16 of 29
1
2
3
4
5
6
7
8
9
23.
24.
Thornton, B. P., Vetvicka, V., Pitman, M., Goldman, R. C., and Ross, G. D. (1996)
Analysis of the sugar specificity and molecular location of the betaꢀglucanꢀbinding
lectin site of complement receptor type 3 (CD11b/CD18), J Immunol 156, 1235ꢀ1246.
Malaga, W., Constant, P., Euphrasie, D., Cataldi, A., Daffe, M., Reyrat, J. M., and
Guilhot, C. (2008) Deciphering the genetic bases of the structural diversity of phenolic
glycolipids in strains of the Mycobacterium tuberculosis complex, J Biol Chem 283,
15177ꢀ15184.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
25.
26.
Fujiwara, T., Hunter, S. W., Cho, S. N., Aspinall, G. O., and Brennan, P. J. (1984)
Chemical synthesis and serology of disaccharides and trisaccharides of phenolic
glycolipid antigens from the leprosy bacillus and preparation of a disaccharide protein
conjugate for serodiagnosis of leprosy, Infect Immun 43, 245ꢀ252.
Chatterjee, D., Cho, S. N., Stewart, C., Douglas, J. T., Fujiwara, T., and Brennan, P.
J. (1988) Synthesis and immunoreactivity of neoglycoproteins containing the
trisaccharide unit of phenolic glycolipid I of Mycobacterium leprae, Carbohydr Res
183, 241ꢀ260.
27.
28.
29.
30.
31.
Fujiwara, T., and Izumi, S. (1987) Synthesis of the neoglycoconjugates of the
phenolic glycolipidꢀrelated trisaccharides for the serodiagnosis of leprosy, Agric Biol
Chem 51, 2539ꢀ2547.
Veeneman, G. H., Van Leeuwen, S. H., Zuurmond, H., and Van Boom, J. H. (1990)
Synthesis of carbohydrateꢀantigenic structures of Mycobacterium tuberculosis using
an iodonium ion promoted glycosidaton approach, J Carbohydr Chem 9, 783ꢀ796.
Fujiwara, T. (1991) Synthesis of the trisaccharideꢀprotein conjugate of the phenolic
glycolipid of Mycobacterium tuberculosis for the serodiagnosis of tuberculosis, Agric
Biol Chem 55, 2123ꢀ2128.
Bourke, J., Brereton, C. F., Gordon, S. V., Lavelle, E. C., and Scanlan, E. M. (2014)
The synthesis and biological evaluation of mycobacterial pꢀhydroxybenzoic acid
derivatives (pꢀHBADs), Org Biomol Chem 12, 1114ꢀ1123.
Barroso, S., Castelli, R., Baggelaar, M. P., Geerdink, D., ter Horst, B., CasasꢀArce,
E., Overkleeft, H. S., van der Marel, G. A., Codee, J. D., and Minnaard, A. J. (2012)
Total synthesis of the triglycosyl phenolic glycolipid PGLꢀtb1 from Mycobacterium
tuberculosis, Angew Chem Int Ed Engl 51, 11774ꢀ11777.
32.
33.
Wei, L., Wei, G., Zhang, H., Wang, P. G., and Du, Y. (2005) Synthesis of new, potent
avermectinꢀlike insecticidal agents, Carbohydr Res 340, 1583ꢀ1590.
Wu, X., MarinoꢀAlbernas, J. R., Auzanneau, F. I., VerezꢀBencomo, V., and Pinto, B.
M. (1998) Synthesis and NMR analysis of 13Cꢀlabeled oligosaccharides
corresponding to the major glycolipid from Mycobacterium leprae, Carbohydr Res
306, 493ꢀ503.
15
ACS Paragon Plus Environment

Page 17 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
34.
35.
David, S., Thieffry, A., and Veyrières, A. (1981) A mild procedure for the regiospecific
benzylation and allylation of polyhydroxyꢀcompounds via their stannylene in nonꢀpolar
solvents, J Chem Soc Perkin 1, 1796ꢀ1801.
Chatterjee, D., Cho, S. N., Brennan, P. J., and Aspinall, G. O. (1986) Chemical
synthesis and seroreactivity of Oꢀ(3,6ꢀdiꢀOꢀmethylꢀbetaꢀDꢀglucopyranosyl)ꢀ(1ꢀꢀꢀꢀ4)ꢀOꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(2,3ꢀdiꢀOꢀmethylꢀ
alphaꢀLꢀrhamnopyranosyl)ꢀ(1ꢀꢀꢀꢀ9)ꢀoxynonanoylꢀbovine
serum
albuminꢀꢀthe leprosyꢀspecific, natural disaccharideꢀoctylꢀneoglycoprotein, Carbohydr
Res 156, 39ꢀ56.
36.
37.
Gurjar, M. K., and Saha, U. K. (1993) Sinthesis of the complete glycotetrapeptide
core of structurally variant Mycobacterium fortuitum glycopeptidolipid, Bioorg Med
Chem Lett 3, 697ꢀ702.
Halcomb, R. L., Boyer, S. H., and Danishefsky, S. J. (1992) Synthesis of the
calicheamicin aryltetrasaccharide domain bearing a reducing terminus: coupling of
fully synthetic aglycone and carbohydrate domains by the Schmidt reaction, Angew
Chem Int Ed Engl 31, 338ꢀ340.
38.
39.
Bundle, D. R., and Josephson, S. (1979) Artificial carbohydrate antigens: synthesis of
rhamnose disaccharides common to Shigella flexneri Oꢀantigen determinants, Can J
Chem 57, 662ꢀ668.
CastroꢀPalomino, J. C., HernandezꢀRensoli, M., and VerezꢀBencomo, V. (1996)
Synthesis of the trisaccharide aꢀLꢀRhaꢀ(1ꢀ2)ꢀaꢀLꢀRhaꢀ(1ꢀ2)ꢀaꢀLꢀRha with a dioxolaneꢀ
type spacerꢀarm, J Carbohydr Chem 15, 137ꢀ146.
40.
41.
Abram, C. L., and Lowell, C. A. (2009) The ins and outs of leukocyte integrin
signaling, Annu Rev Immunol 27, 339ꢀ362.
Shi, Y., Tohyama, Y., Kadono, T., He, J., Miah, S. M., Hazama, R., Tanaka, C.,
Tohyama, K., and Yamamura, H. (2006) Proteinꢀtyrosine kinase Syk is required for
pathogen engulfment in complementꢀmediated phagocytosis, Blood 107, 4554ꢀ4562.
Troegeler, A., Lastrucci, C., Duval, C., Tanne, A., Cougoule, C., MaridonneauꢀParini,
I., Neyrolles, O., and LugoꢀVillarino, G. (2014) An efficient siRNAꢀmediated gene
silencing in primary human monocytes, dendritic cells and macrophages, Immnol Cell
Biol 92, 699ꢀ708.
42.
43.
Cambier, C. J., Takaki, K. K., Larson, R. P., Hernandez, R. E., Tobin, D. M., Urdahl,
K. B., Cosma, C. L., and Ramakrishnan, L. (2014) Mycobacteria manipulate
macrophage recruitment through coordinated use of membrane lipids, Nature 505,
218ꢀ222.
16
ACS Paragon Plus Environment

ACS Chemical Biology
Page 18 of 29
1
2
3
Supporting Information Available: This material is available free of charge on the ACS
4
5
6
website. Procedures for the synthesis of PGL epitopes: general procedures and preparation
1
of individual compounds. Physicoꢀchemical data for new compounds: optical rotations, H
7
8
9
and 13C NMR data, high resolution mass spectra data. Copies of 1H and 13C NMR spectra for
1
new compounds. NMR analysis of PGL purified from rBCG ꢀRv2959:: ML0126. Copy of H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NMR spectra for DIM and PGL purified from rBCG ꢀRv2959:: ML0126. Methods, references
and supplementary figures.
17
ACS Paragon Plus Environment

Page 19 of 29
ACS Chemical Biology
1
2
3
FIGURE LEGENDS
4
5
6
7
8
Figure 1. Structures of PGL produced by various mycobacterial pathogens and
construction of the rBCG strains. Structure of the major PGL lipid core (DIM) and the
speciesꢀspecific sugar moieties of the major forms of PGL produced by M. tuberculosis
(PGLꢀtb), M. ulcerans (PGLꢀulc) and M. leprae (PGLꢀ1). Genes coding for the enzymes
required for the genetic reprogramming of M. bovis BCG are indicated.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Analysis of recombinant BCG strains. (A) TLC analysis of lipids extracts from
various rBCG producing or not GFP. (B) TLC analysis of glycolipids from rBCG ꢀRv2959 and
rBCG ꢀRv2959 expressing ML0126.
Figure 3. Elaboration of the glycoside building blocks
Figure 4. Synthesis of PGL saccharide epitopes (OS PGLs)
Figure 5. PGL-1 is unique to promote human macrophage invasion via the lectin
domain of CR3. (AꢀB) Percentage of infected hMDM after 2h of contact with PGLꢀ
expressing rBCG strains at MOI 5:1 was quantified by fluorescence microscopy. (B) Effect of
the preꢀincubation with 10 ꢁg mlꢀ1 blocking antiꢀCR3 antibodies 2LPM19c or VIM12, or
isotype control, on the percentage of infected hMDM. Data are expressed as mean ± SEM
and are representative of three independent experiments performed in duplicate. * P < 0.05,
** P < 0.01
Figure 6. Lyn is required for PGL-1-dependent increase in hMDM uptake. hMDM were
transfected with control siRNA or siRNAs targeting Lyn or Syk. 96h after scrambled or siRNA
transfection, cell lysates were subjected to Western blot using specific antibodies to assess
Lyn and Syk knockdown (A). hMDM were infected with BCG and rBCGꢀPGLꢀ1 strains at MOI
5:1 for 2h and (B) the percentage of infected hMDM was quantified. Data are expressed as
mean ± SEM and are representative of three independent experiments performed in
duplicate. ** P < 0.01
Figure 7. Binding of PGL-1 to CR3 is mediated by its saccharide domain. (A) Increasing
concentrations of purified PGLꢀ1 (B), or purified PGLꢀ1 (5 ꢁM) or its synthetic oligosaccharide
moiety (OSꢀPGLꢀ1) (50 ꢁM), were incubated with immobilized human CR3. Bound PGLꢀ1
was detected using an antiꢀPGLꢀ1 antibody. (C) Purified CR3 was preꢀincubated, or not, with
18
ACS Paragon Plus Environment

ACS Chemical Biology
Page 20 of 29
1
2
3
4
5
the synthetic oligosaccharides (OSꢀPGLs) before being exposed to coated PGLꢀ1. Bound
CR3 was detected using an antiꢀCD11b antibody. Data are expressed as mean ± SEM and
are representative of at least 3 independent experiments performed in triplicate. ** P < 0.01.
6
7
8
9
Figure 8. PGL-1 and PGL-tb inhibit TLR-2-induced NF-κB activation and TNF-α
secretion. NFꢀκBꢀreporter THPꢀ1 cells were infected with PGLꢀexpressing rBCG strains at
MOI 1:1 (A) or treated with 5 ꢁM PGLꢀ1, 10 ꢁM DIM or 50 ꢁM synthetic oligosaccharides
(OSꢀPGLs) in the presence of the TLR2 agonist Pam3CSK4 (10 ng mlꢀ1) (C), and
phosphatase activity was quantified after 24h. hMDM, either untreated (B) or transfected with
control siRNA or siRNAs targeting Lyn or Syk for 96h (F), were infected with PGLꢀexpressing
rBCG strains at MOI 10:1 for 2h and the secretion of TNFꢀα was quantified. (D) HEKꢀTLR2
NFꢀκBꢀreporter cells were treated with the lipids or the OSꢀPGLs in the presence of
Pam3CSK4 at the concentrations indicated above, and phosphatase activity was quantified
after 24h. (E) 5 ꢁM purified PGLꢀ1 was incubated with immobilized human TLR2. Bound
PGLꢀ1 was detected using an antiꢀPGLꢀ1 antibody. (AꢀD,F) Histograms represent the mean ±
SEM of at least 3 independent experiments performed in triplicate. * P < 0.05; ** P < 0.01;
*** P < 0.001, or (E) are representative of 3 independent experiments performed in triplicate.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
19
ACS Paragon Plus Environment

Page 21 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Structures of PGL produced by various mycobacterial pathogens and construction of the rBCG
strains. Structure of the major PGL lipid core (DIM) and the species-specific sugar moieties of the major
forms of PGL produced by M. tuberculosis (PGL-tb), M. ulcerans (PGL-ulc) and M. leprae (PGL-1). Genes
coding for the enzymes required for the genetic reprogramming of M. bovis BCG are indicated.
62x28mm (300 x 300 DPI)
ACS Paragon Plus Environment

ACS Chemical Biology
Page 22 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Analysis of recombinant BCG strains. (A) TLC analysis of lipids extracts from various rBCG
producing or not GFP. (B) TLC analysis of glycolipids from rBCG ∆Rv2959 and rBCG ∆Rv2959 expressing
ML0126.
96x67mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 23 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Elaboration of the glycoside building blocks
183x250mm (300 x 300 DPI)
ACS Paragon Plus Environment

ACS Chemical Biology
Page 24 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Synthesis of PGL saccharide epitopes (OS PGLs)
221x400mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 25 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. PGL-1 is unique to promote human macrophage invasion via the lectin domain of CR3. (A-B)
Percentage of infected hMDM after 2h of contact with PGL-expressing rBCG strains at MOI 5:1 was
quantified by fluorescence microscopy. (B) Effect of the pre-incubation with 10 µg ml-1 blocking anti-CR3
antibodies 2LPM19c or VIM12, or isotype control, on the percentage of infected hMDM. Data are expressed
as mean ± SEM and are representative of three independent experiments performed in duplicate. * P <
0.05, ** P < 0.01
101x177mm (300 x 300 DPI)
ACS Paragon Plus Environment

ACS Chemical Biology
Page 26 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Lyn is required for PGL-1-dependent increase in hMDM uptake. hMDM were transfected with
control siRNA or siRNAs targeting Lyn or Syk. 96h after scrambled or siRNA transfection, cell lysates were
subjected to Western blot using specific antibodies to assess Lyn and Syk knockdown (A). hMDM were
infected with BCG and rBCG-PGL-1 strains at MOI 5:1 for 2h and (B) the percentage of infected hMDM was
quantified. Data are expressed as mean ± SEM and are representative of three independent experiments
performed in duplicate. ** P < 0.01
81x113mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 27 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Binding of PGL-1 to CR3 is mediated by its saccharide domain. (A) Increasing concentrations of
purified PGL-1 (B), or purified PGL-1 (5 µM) or its synthetic oligosaccharide moiety (OS-PGL-1) (50 µM),
were incubated with immobilized human CR3. Bound PGL-1 was detected using an anti-PGL-1 antibody. (C)
Purified CR3 was pre-incubated, or not, with the synthetic oligosaccharides (OS-PGLs) before being exposed
to coated PGL-1. Bound CR3 was detected using an anti-CD11b antibody. Data are expressed as mean ±
SEM and are representative of at least 3 independent experiments performed in triplicate. ** P < 0.01.
52x19mm (300 x 300 DPI)
ACS Paragon Plus Environment

ACS Chemical Biology
Page 28 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Both PGL-1 and PGL-tb inhibit TLR-2-induced NF-κB activation and TNF-α secretion. NF-κB-reporter
THP-1 cells were infected with PGL-expressing rBCG strains at MOI 1:1 (A) or treated with 5 ꢀM PGL-1, 10
ꢀM DIM or 50 ꢀM synthetic oligosaccharides (OS-PGLs) in the presence of the TLR2 agonist Pam3CSK4 (10
ng ml-1) (C), and SEAP activity was quantified after 24h. hMDM, either untreated (B) or transfected with
control siRNA or siRNAs targeting Lyn or Syk for 96h (F), were infected with PGL-expressing rBCG strains at
MOI 10:1 for 2h and the secretion of TNF-α was quantified. (D) HEK-TLR2 NF-κB-reporter cells were treated
with the lipids or the OS-PGLs in the presence of Pam3CSK4 at the concentrations indicated above, and
SEAP activity was quantified after 24h. (E) 5 ꢀM purified PGL-1 was incubated with immobilized human
TLR2. Bound PGL-1 was detected using an anti-PGL-1 antibody. (A-D,F) Histograms represent the mean ±
SEM of at least 3 independent experiments performed in triplicate. * P < 0.05; ** P < 0.01; *** P < 0.001.
(E) Data are expressed as mean ± SEM and are representative of 3 independent experiments performed in
triplicate.
101x75mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 29 of 29
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TOC graphic
80x44mm (300 x 300 DPI)
ACS Paragon Plus Environment